Controles de saldos de operaciones y garantías (C02)

Ethical clearance to conduct the study was sought from MUHAS Ethical Review Board.

Permission to do the study was obtained from MNH and Psychiatry Head of Department.

CHAPTER THREE 3.1 RESULTS

3.1.1 Univariate Analyses

3.1.1.1 Description of participants

A total of 632 participants‟ files enrolled at MNH MAT clinic between February 2011 and February 2013. Data was not extracted from files of 23 (3.64%) participants who had passed away prior to February 2013. Final analysis included data from 609 participants‟

files (figure 1). Male participants were in the majority compared to female participants.

They comprised 93.3% (568) of total population. The mean age (SD) was 34.28 years (6.41).

Table 1 below, shows the distribution of the study participants‟ characteristics (socio-demographic, health status and psychosocial and behavioral). Results of the current study show that with regard to the socio-demographic characteristics a high number of participants reported to be single 480 (78.8%) compared to those who were married 81 (13.3%) and widowed or divorced 44 (7.2%). Regarding level of education, most study participants attending MAT clinic completed primary school education 294 (48.3%) followed by those who had secondary education (161, 26.4%). The majority of participants 462 (75.9%) reported to have no employment, followed by 7.6% who had part time employment. Furthermore, among 609 participants attending MAT clinic 42.7% reported illegal activities to be their source of income. Finding related to their health status shows that 277 (45.5%) have never been hospitalized, 199 (32.7%) had a low number of hospitalization and 133 (22 %) reported a high number of hospitalization. A total of 76 (12.5%) participants reported to have chronic illnesses that impaired their functioning compared to 74.9% who stated not to have a chronic illness. Among 454 participants who were tested for HIV at the clinic, 170 (27.9%) tested positive and 284 (46.6%) tested negative. The serostatus of the remainder of clients was unknown, at least not recorded on file. On the Hopkins Symptom checklist for anxiety and depression 389 (63.9%) reported low scores while 209 (34.3%) scored high on the symptoms.

The majority of the participants scored low on depressive symptoms as compared to those who scored high (73.1% versus 3.8%). Participants‟ substance use dependence status was also assessed and 98 % met the criteria for dependence only 2 (0.3%) participants did not meet criteria for dependence. Dependence is one of the criteria for enrolment it is not clear how two clients who do not meet the criteria are attending, data collected from baseline

1-2 Dependants 174 (28.6)

Incarceration²

Never 341 (56.0)

Once 65 (10.6)

More than once 169 (27.8)

Anxiety symptoms

Low 389 (63.9)

High 209 (34.3)

Depression Symptoms

Low 445 (73.1)

High 23 (3.8)

1All variable totals < 609 are due to missing responses

The psychosocial factors of participants explored included the quality of life (QOL).

Results show that a high proportion (46.8%) of participants were found to have poor quality of life, while 25.5% had average, 14.6% and 4.8% had good and very good quality of life respectively. On criminal background, 341 (56%) have never been in jail, 65 (10.6%) only once and 169 (27.8%) more than once. With regard to living arrangement, a high proportion of participants stay with their parents (29.6%) (See Table 1)

3.1.1.2 Description of methadone treatment adherence

As seen in Figure 1, the proportion of participants who adhered to the methadone treatment among IDUs attending MAT clinic at MNH was 75% (460) and 25% (149) did not adhere to the treatment.

Figure 3: Adherence

Adhered (n=460) 75%

Dropout (n=149) 25%

Figure 2. Adherence

3.1.2 Bivariate Analyses

3.1.2.1 Factors associated with methadone treatment adherence

As evident from Table 2 below, there was a significant association between adherence and employment as source of income (p=0.031), observing crudes odds ratio as summarized in Table 3, results shows that participants whose source of income is from employment are 1.5 times more likely to adhere to methadone treatment (OR, 1.50, 95% CI: 1.00-2.23) compared to those who are not employed. The association between adherence and other social demographic factors (age, gender, marital status and level of education) was not significant. Hospitalization was another factor that appears to be significantly associated with adherence to methadone treatment (p=0.027). Crudes odds ratio suggest that participants with a low number of hospitalization compared to no hospitalization are 49%

less likely to adhere (OR, 0.51, 95% CI: 0.29-0.86) and those who have a high number of hospitalization compared to no hospitalization were also less likely to adhere to methadone treatment (OR, 0.50, 95% CI: 0.29-0.87). Levels of anxiety, depression or HIV status were not significantly associated with adherence. However, incarceration had a significant relationship with adherence (p=0.017) with crudes OR of 1.71 (95% CI: 1.08-2.72). There was also a significant relationship between quality of life and adherence (p=0.028).

Table 2. Bivariate analysis of predictors of adherence to MAT

Hospitalization² 7.26 0.027

Low (1-4) 49 (26.1) 139 (73.9) 188 (100) High (above 90) 47 (22.1) 166 (77.9) 213 (100)

Life time incarceration² 8.15 0.017

Never 92 (27.0) 249 (73.0) 341 (100)

Once 22 (33.8) 43 (66.2) 65 (100)

More than once 30 (17.8) 139 (82.2) 169 (100)

Dependence 1.150 0.358

1-4 1 (50) 1 (50) 2 (100)

5-7 92 (19.7) 375 (80.3) 467 (100)

Anxiety symptoms 2.10 0.349

Low 129 (26.1) 366 (73.9) 495 (100)

Moderate 16 (18.8) 69 (81.2) 85 (100)

High 4 (22.2) 14 (77.8) 18 (100)

Depression Symptoms 2.05 0.166

Low 146 (24.6) 447 (75.4) 593 (100)

High 3 (50) 3 (50) 6 (100)

1 All variable totals < 609 are due to missing responses

2 p<0.05

3.1.3 Multivariate Analyses

3.1.3.1 Factor associated with adherence to methadone treatment

All social-demographic, psychosocial and health related factors that had p<0.25 were explored in the forward logistic model to determine which factor best predicts the likelihood of adherence to methadone treatment among IDUs at MNH methadone clinic.

Gender was the only factor that was found to show a significant association with adherence. Male participants compared to female were 0.24 less likely to adhere to treatment (OR, 0.24, 95% CI: 0.07-0.85). (Table 3)

Table 3. Logistic regression of independent predictors of adherence to MAT among IDUs in Dar es Salaam, Tanzania

1N=609 (%) Adherence

Average 155 (25.5) 0.42

1All variable totals < 609 are due to missing responses

2 p<0.05

3 Adjusted for all variables in the table

CHAPTER FOUR 4.1 DISCUSSION

This study aimed at identifying individual factors that are associated with treatment adherence among IDUs attending Methadone Assisted Therapy (MAT) at Muhimbili National Hospital (MNH). The findings of the study provide a descriptive distribution of characteristics of clients who attend the clinic and the factors that influence whether they remain in treatment or drop out.

In answering the first research question of the study “What is the proportion of IDUs adhering to methadone assisted therapy at MNH? Clinical effectiveness of methadone is most commonly measured by retention of clients in treatment and by abstaining or reduction in the frequency of concurrent drug use (heroin and /or others) compared to no treatment or treatments that do not include methadone (Gowing et al., 2008, Mattick et al., 2009). Adherence in this study is operationalized as length of attendance at MAT in days, non adherence means absent for 30 days or more continuously after which returning back into programme needs NGO intervention using patient‟s Case Worker.

Based on this definition the study shows that 75% of MAT clients adhered in that they were retained in MAT at time of study (over a two year period from time of enrolment Feb 2011 to Feb 2013) and 25 % dropped out with that period and prior to completion of treatment .

The retention rate is higher when compared to Liu et al., (2008) reported results from a prospective cohort study in Guizhou province, China which recruited 1003 patients from 8 MAT clinics. In this study, the overall retention rate after 14 months of follow-up was 56.2%. It was found that a higher dose of methadone predicted higher retention.

Also findings from this study are higher that a retention rate of 48.7 % (drop up rate of 51.3%) revealed in a more recent study conducted in China by Gu et al., (2012) among 158 methadone assisted therapy clients. And even higher when compared to a study in Taiwan that reported a retention rate of 32.3% (drop out 67.7%) among 368 IDUs attending a methadone clinic during 18 months follow up (Lin et al., 2013). Barriers to

adherence included: side effects, inconvenience, low dosage, financial difficulties, and lack of access.

The difference in adherence between Tanzania and countries like China could reflect the difference in culture as well as programme related factors.

However an analysis of a 9-month follow-up data from a cohort of 965 patients in a pilot project in Vietnam revealed that the treatment retention rate after 9 months was approximately 90% after one year and 80% after two years (Long et al., 2010) which compares favourably with the findings of this study. Data which is available and important in understanding the factors associated with adherence but not reported in this study include: point in time of drop out (6, 9, 12, months), reasons for drop out, proportion of clients with concurrent drug use (heroin or other drugs) or to what degree they have reduced heroin use (either self-reported or detected by urine testing).

One factor that might have contributed to the success in terms of retention in treatment is the involvement of a family member (adherence sponsor) in the role of the treatment supporter as a requirement for every patient both in Dar es Salaam and Vietnam.

Additionally, both are the pilot programmes conducted with high level of attention and dedicated resources from many sectors of the society.

Among factors that contributed to the higher adherence rate at MAT Clinic at MNH are;

Before clients are enrolled to the Methadone program at MNH they have to be screened and make a contract with one of the selected NGOs. The NGO provide follow up of clients as well as support. Clients at Methadone Clinic at Muhimbili, they have to maintain a good attendance, good behavoir. If they misbehave, they are first given a verbal warning, then written warning. If they still misbehave, they are sent back to the NGO where they came from, where they are counseled and they will be sent back the Methadone Clinic. At the clinic, beside having Psychiatrists, Pharmacists , and Nurses, there are also, Clinical Psychologists, Social workers, Occupational Workers who strengthen the services. Clients can participate in individual or group therapy. There is a program whereby clients after showing good manner are being employed to work as gardeners at the MNH campus and get paid

The second research question of the study asks “Is there are any associations between social demographic factors and adherence to methadone treatment among IDUs?” Finding of the current study shows that when adjusted for predictors of methadone treatment adherence, gender was the only social demographic factor that showed a significant association. Male IDUs attending MAT compared to female were 76% less likely to adhere (AOR, 0.24, 95% CI: 0.07-0.85). While there is some support in the literature to support this association it is not very robust. And the fact that males accounted for 93.3% of the sample and females only 6.7% reduces the strength of the association. However, Agosti et al (1991) found that males were more likely than females to drop out of treatment, and other studies have supported this showning females to be more likely to adhere in treatment as well (Chou et al., 1998; Simpson et al., 1997).

However, there are studies that have reported that women were more likely to drop out of treatment (Arfken et al., 2002; Hser et al., 2004; Roberts & Nishimoto, 1996) and in outpatient treatment patients, men remained in treatment longer than women (McCaul et al.,2001). Mammo et al., (1993) in a study involving 2,697 outpatients and intensive outpatient admissions for the state of New Jersey found that the rate of not completing treatment was significantly higher for females, those who are less educated, those employed in less – skilled occupation and the young. And yet other studies have found that gender has not been related to treatment success or failure (Carroll et al.,1991; Gainey et al., 1993; Melnick et al., 1997; Saum et al., 2001).

For non adjusted factors (predictors), there was an interesting finding which suggest that methadone clients who do not have family members who depend on them are less likely to adhere to methadone treatment compared to those who have more than 3 dependents.

Comiskey C.M (2013) in a 3 year national longitudinal study comparing drug treatment outcomes for opioid users with and without children in their custodial care at intake demonstrate that having children in a client's care improves outcomes for heroin use but also suggest the possible use of substitution substances.The association could imply that those who do not have people who depend on them are less likely to be motivated to recover, while those who need to support their family members may feel the need to stay in the program and be able to provide the requred support

In relation to other social demographic factors, findings of the current study are supported by McHugh et al., (2013) who found marital status, education and employment not to be associated with methadone treatment adherence. However, other previous studies (Chou et al., 1998; Hser et al., 2004; Veach et al, 2000; Mammo et al., 1993) do not support the current studies on the identified demographic factors.

HIV prevalence rate of 25% was noted in this study, which is lower than estimated by others.

HIV status for 155 clients was unknown at baseline. It is not clear whether clients did not know or they refused to reveal their status. From the community studies Williams et al., (2009) looked at HIV prevalence in sexually active IDUs in Dar es Salaam using serology testing after they consented to giving a blood sample. 64% of the females were found to be HIV positive and 28% of the men despite only 5 participants saying they were positive in the pre-test interview. In order to provide educational intervention programs for IDU in terms of lowering HIV risk, more knowledge of IDU beliefs and attitudes regarding HIV risk is needed for 6 and 12 months follow up.

The third research question aimed at exploring the association between psychosocial factors including quality of life as determined by scores on the SF 12 (HRQOL) , social support, living arrangements and criminal activities and adherence to methadone treatment.

When adjusted, there was no significant association between psychosocial factors and methadone adherence.

Results of the current study show that those who have poor quality of life compared to those who have very good are less likely to adhere and this is consistent with findings from Mc Lellan et al., (1983). Findings from this study do not suggest a relationshp between living arrangemets, social support, criminal activity and adhereance. These findings are not consistent with other studies (Sayre et al., 2002; Tuten et al., 2003; Broome et al., 2002) who found social relationship, living arrangement, social support, quality of life and criminal background to be associated with methadone adhrence.

However, on the bivariate analysis, criminal background suggest that those who have spent more time in jail are more likely to adhere. This is in contrast to Waldorf et al., (1983) who found that those with criminal background are less likely to adhere to treatment. Hiller et al., (1985) examined the association between legal pressure and treatment adherence in a national sample of 2,605 clients admitted to 18 long-term residential facilities that participated in the Drug Abuse Treatment Outcome Study (DATOS). They found that those who entered residential treatment with moderate to high pressure from legal authorities adhered less compared to those who entered under low pressure from the legal authorities.

Mattick and colleagues (2009) meta analysis on the effectiveness of MAT reported that MAT may reduce criminal activity even though the summary effect estimate is not statistically significant and indicated that the non-significant results for criminal activity reductions may be due to scarce data on these two rather uncommon outcome measures of effects of MAT.

A pilot study in Vietnam reported significant reduction in HIV-related risk behaviours among patients and in crime activities in the districts where the pilot MAT was implemented; and significant improvement in patient quality of life (Long et al., 2010).

The fourth research question asks “Does HIV/AIDS status influence adherence to methadone treatment among IDUs” Results of this study showed that HIV/AIDS status has no significant relationship with methadone treatment adherence. It is well established that transmission of HIV can be reduced among opiate addicts by making drug treatment available to those that want help. Harris et al., (2006) identified a relative low rate of HIV infection among MAT patients. They suggested a possible protective effect of methadone maintenance against acquiring HIV infection, because MAT is orally administered and is not injected (which is a primary route for HIV infection among drug users).

However evidence to suggest that HIV is associated with adherence is limited. McHugh et al., (2013) did not find an association when examining 50 IDUs living with HIV attending methadone. Sendi et al., (2003) have shown that HIV negative status has not been associated with adherence.

With regard to the number of hospitalization and its relationship to treatment retention, the findings shows that participants with more frequent hospitalizations are less likely to adhere to the treatment compared to those who have not been admitted at the hospital.

Studies have shown that older opiate dependent individuals who continued using heroin had higher rates of disability and more frequent hospitilizations (Hser et al., 2001). It would be interesting to note if there is a relationship between freuquency of hospitilizations and age and stage of HIV infection in this study

The last reseach question was to identify whether mental conditions (anxiety and depresion) are associated with methadone treatment adherence. Findings from this showed insignificant association between anxiety or depression and methadone treatment adherence. Results are inconsistent with findings reported by Gonzalez et al., (2011) who found that depressed patients are three-times more likely to be non adherent as non depressed patients. The lack of association in this study could be accounted for in part in that the scores on the John Hopkins depression was not completed in 141 files.

Amodeo et al., (2008) showed association between anxiety and adherence to methadone treatment. The lack of association again could reflect the limitation of the deign. Even though the scores for anxiety were available for all but one client, anxiety (and depression) were recorded during the initial assessment by rotating clinicians at the clients intake assessment, whereas adherance in terms of retention or drop out as it occurred later at 6 or 12 months follow up. Initial anxiety and depressed symptoms may hae resoled by time of

„drop‟out‟. To examine these mental condition there is a need for a longitudinal case control design that will allow for a better comparison.

4.2 STUDY LIMITATIONS

There are a number of factors that may have affected the quality of the study. One of the main limitation was the design of the study. The principal investigator relied primarily on the electronic recorded data. The study depended on the information in the patient‟s files.

Secondary data collection may have high rates of missing data as it completely depends on data collected before the study.

For example adherence in the context of MAT services, includes not only attendance and following the treatment regime but also abstinance or reduction in continued use of heroin or other drugs as determined by self report or urine testing. Only data on attendance and drop out were retrieved for analysis which is only one aspect of adherence. Also it should have been possible to correlate point of drop out and lenght of time in the programme.

This would have given some important data on pattern of drop out (e.g early in the programme or later- 6, 12, 18, 24 months).

This is the first time for clients have had an opportunity to access treatment and the enrollment procedures are strict. Clients may have felt the need to respond in a social desirable way because of fear of non-enrollment or termination from the program.

In addition the current study did not include data on a number of additional treatment adherence influencing factors. These include program related reasons for drop out, psychological factors like self-efficacy, coping and stress, dose associated with retention, dose and HIV status, satisfaction with the programme, structural issues like distance from the site, unable to leave the city because of daily attendance.

CHAPTER FIVE 5.1 CONCLUSION AND RECOMMENDATION

It is evident from the relativly low drop out rate (25% at two year follow up) that factors that are associated with non adherence are been addressed to a great extent in the MAT

It is evident from the relativly low drop out rate (25% at two year follow up) that factors that are associated with non adherence are been addressed to a great extent in the MAT